Abstract: In heavy oil, heavy components form aggregates through various intermolecular interactions, which contribute significantly to its high viscosity. However, there is currently a lack of quantitative studies on the contribution of these interactions to the viscosity of heavy oil. Based on the typical average molecular structure of heavy oil from the Block Zheng 364 of Shengli Oilfield, a molecular component model was constructed using the modified Boduszynski-Li (B-L) method. Combined with molecular dynamics simulations and the “component extraction method,” the influence of changes in mass fraction of heavy components on the system’s viscosity was systematically investigated. Principal component analysis (PCA) was employed to quantitatively analyze the intrinsic correlations between macroscopic viscosity and intermolecular interactions, including van der Waals, Coulombic, hydrogen bonding, and π-π stacking, etc. Simulation results indicate that asphaltenes and resins are the key components responsible for the high viscosity of heavy oil. It is clarified that the contribution of resins and asphaltenes to viscosity primarily originate from π-π stacking and hydrogen bonding, respectively. Hydrogen bonding interactions between components contribute most significantly to the system’s viscosity in the heavy oil of the Block Zheng 364. By coupling the “component extraction method” with PCA statistical analysis, this study quantitatively unveils the contribution of heavy oil components and their interactions to viscosity at the molecular scale, thereby deepening the understanding of the contribution mechanism in heavy oil viscosity. The findings provide a theoretical basis for the screening of high-efficiency viscosity reducers.